Rehearse a depot outage before enabling live dispatch
Dispatch needs to know what remains usable when telemetry stops but assignment storage still works. Write a fixed-data outage rehearsal and an operator decision note.
- Focused work estimate
- 2h + prerequisites
- Priority in the scenario
- High
- Engineering practice
- Failure testing · Operational handoff · Observability
Estimated field mix
- Real-time systems40%
- Site reliability30%
- Quality engineering30%
Field percentages are editorial estimates of the ticket's engineering focus. They total 100%; they are not measured time, proficiency scores, or ownership evidence.
Review it, then add it to your workspace.
The board opens an editable draft; nothing is saved until you confirm it. Sign-in and workspace permissions apply, and Demo boards remain ephemeral.
Project context
The fictional Cedar delivery depot dispatches vans through a browser map and accessible list. Synthetic telemetry arrives through a gateway with device sequence numbers. This is operational fleet software practice; no real driver tracking, employee scoring, or routing safety guarantee is part of the project.
Setup prerequisites
- Synthetic vehicle telemetry with device sessions and sequence numbers
- Dispatch assignment and authorized region API contracts
Preceding work
Complete these dependencies, or supply their agreed outputs before taking this ticket.
- FLEET-101 · Keep invalid coordinates from appearing as real vans
- FLEET-103 · Fit selected vehicles without zooming out across the entire planet
- FLEET-102 · Show dispatchers how old each vehicle position is
- FLEET-104 · Prevent delayed telemetry from moving a van backwards
- FLEET-105 · Reconnect the map without losing updates between snapshot and stream
- FLEET-106 · Resolve competing dispatch assignments explicitly
- FLEET-108 · Remove out-of-region location data when dispatch scope changes
- FLEET-107 · Keep map, list, and counts on the same filtered vehicle set
- FLEET-109 · Coalesce position rendering without dropping dispatch events
Acceptance criteria
- The rehearsal separates telemetry outage, assignment-store outage, and authorization outage.
- Every scenario records visible freshness, permitted actions, and recovery state.
- The note defines safe fallback actions and states that synthetic replay is not proof of real fleet readiness.
Implementation constraints
- Do not send real dispatch commands or connect to personal/production location feeds.
Verification to include
- Stop telemetry, advance the clock, and recover by snapshot; stale labels and final positions match the trace.
- Fail assignment storage and then authorization during a pending command; show no false success and close delivery when scope is unknown.
Deliverables
- Depot outage rehearsal and dispatch operator runbook
Rollout and recovery
Require a reviewed rehearsal for pilot use; pause live dispatch when authority or assignment confirmation is unavailable.
Value of the work
For the engineer: Practice geospatial boundaries, stream ordering, snapshot recovery, and concurrent operational commands.
For the team: Inspect whether a developer keeps dispatch decisions grounded in current authorized data and exposes uncertainty during outages.
Evidence boundaries
Outcome Evidence: Tests, patches, and runbooks are requested deliverables. They become Outcome Evidence only through a qualified Mission and immutable Evidence IDs.
Ownership Evidence: Independent adaptation must be observed under a declared verification policy and cite immutable Evidence IDs. Completing a planning ticket establishes no Ownership Evidence.